The Basics of Revenue Management

I'm taking this great class at SMU this semester called revenue management. It is a very math intensive class and involves a lot of statistical modeling in Excel. We are using a relatively new concept called Dynamic Programming which is a methodology to solve complex optimization problems over time.

While these theories aren't necessarily new, they do not have widespread adoption in the business world yet. There are only 3 MBA programs in the country teaching it right now, and only two of them teach it for credit. This class is mostly seen just at the PhD level right now, but that is going to change.

Here's how it works:

Lets say you are the Red Sox (I can always bring my examples back to baseball). You've just aquired a Japanese superstar pitcher and want to maximize your revenue generation opportunities with him. You already sell out your ballpark for the entire year, so selling new seats seem like it won't work for you. However, this isn't the case - you do have the choice to not sell out every game and instead hold back a number of seats to sell at the last minute when you know the superstar will be pitching. But how many do you hold back and how much do you charge?

Here's where Revenue Management comes in. It helps you decide how many seats you should hold back, potentially how much you should charge for them, and when you should sell them. See, the problem is not whether you should hold the seats back and soak the poor tourists who are willing to pay ungodly amounts of money at the last minute to see a particular pitcher. You know you should do that. The problem is how many and for how much. If you try to hold back too many, you risk the possibility that no one buys and you are stuck with a bunch of unsold seats that you could have sold to a million other people 3 months ago.

If this sounds eerily familiar, it should. This model has been used for about 20 years by the airline industry. We all know it as "three week advanced booking." The dirty little secret is that if 150 people show up for a 150 seat plane four week ahead of time, the airline will not sell it out and will hold the seats back for higher revenue customers at the last minute.

The dirtier secret is called over-booking. You might think that over-booking is caused by tickets being sold in multiple places and they just can't coordinate well enough and mistakes happen. On the contrary, overbooking is planned and encouraged. See, the airlines have figured out and have a lot of data on no-shows. They've computed the probabilities and are milking it for every dollar they can. They know that if too many passengers show up that they will have to give them compensation, but they've computed that into the overbooking equation as well.

My professor thinks, and I agree, that this mode of thinking will slowly invade every aspect of startegic thinking in companies over time. As the tools and concepts get easier to use, companies will leverage it.

A manufacturing plant in China that normally books capacity out months or years ahead of time will start holding back capacity for rush orders. More importantly, they will know how much capacity to hold and how much to charge.

Target will hold inventory of the new Elmo-super-duper-tickle-me-with-extreme-giggle-technology back until New Years' eve and charge a hefty premium for it.

You will see this everywhere.